I think there is still work to do in differential geometry/topology aspects of GR.
Some folks still work on aspects of causal structure and singularity theorems.
Some folks would like to find ways of generating solutions of the field equations... and this could include exploiting symmetries and classifying spacetimes according to available structures [e.g. symmetries imposed, choice of matter fields]. There are probably aspects of differential geometry/topology in the "space of solutions".
Some folks are looking to reformulate GR in terms of new variables and new structures which might prove simpler for formulating initial-value problems, for analysis, for computation, for "quantization", or for further generalization (including quantum field theory in curved spacetime and approaches to quantum gravity and unified theories).
Certainly more differential geometric/topological tools (including development of pedagogy) could be helpful in trying to help others extract the physics from it.
Some possibly interesting reading...
"Results and Open Problems in Mathematical General Relativity" - Yvonne Choquet-Bruhat
http://www.springerlink.com/content/964186644455l058/
Singularity Theorems in General Relativity: Achievements and Open Questions - José M.M. Senovilla
http://arxiv.org/abs/physics/0605007
83 years of general relativity and cosmology: progress and problems - George F R Ellis
http://www.iop.org/EJ/abstract/0264-9381/16/12A/303
...
http://www.google.com/search?hl=en&q=open+(problems+OR+questions)+"general+relativity"